Rapid identification of bacteria and prompt acquisition of susceptibility results are valuable for patient care. The objective of the present study was to determine the accuracy of direct inoculation of Vitek 2 cards from positive BACTEC cultures compared to inoculation of the cards from subculture plates. Positive BACTEC cultures sampled between March 2001 and June 2002 were included. The results of direct inoculation were compared with the results of inoculation of Vitek 2 cards from subcultures. Of 161 gram-negative bacilli, 129 (80%) were correctly identified by direct inoculation compared to 145 of 161 (90%) by subculture. Susceptibility testing was performed on 2,862 antibiotic-isolate combinations. The essential agreement was 98.7%. The number of very major, major, and minor errors was 1 (0.2% of resistant strains), 1 (0.04% of susceptible strains), and 68, respectively. Direct identification of Staphylococcus spp. was not performed, but antimicrobial susceptibility was tested using 6,042 antibiotic-isolate combinations. The essential agreement was 95.2%. The number of very major, major, and minor errors was 73 (4.5% of resistant strains), 32 (0.8% of susceptible strains), and 106, respectively. Eighty-four percent of the very major errors occurred with trimethoprim-sulfamethoxazole. The results show that direct inoculation of Vitek cards is valuable as a rapid routine method for identification and susceptibility testing of gram-negative bacilli. For Staphylococcus spp., the susceptibility results obtained after direct inoculation of Vitek 2 cards are also acceptable except for those obtained with trimethoprim-sulfamethoxazole. Susceptibility results for this antibiotic, if obtained using direct inoculation, should not be reported to the clinician.